Go back to Market Research
Blue Yonder supplies software for retailers, manufacturers, distributors, and logistics operators. Its portfolio covers demand and supply planning, replenishment, production scheduling, order promising, warehouse and transportation operations, and returns. Customers use these applications to calculate orders and schedules, allocate inventory, execute fulfillment, and coordinate trading partners. The review covers this portfolio, including its Cognitive Planning and agent offerings. (1, 15, 25)
Overall score: 5.0/10
| Dimension | Score |
|---|---|
| Supply chain depth | 5.4/10 |
| Decision and optimization substance | 6.0/10 |
| Product and architecture integrity | 4.2/10 |
| Technical transparency | 5.0/10 |
| Vendor seriousness | 4.4/10 |
Blue Yonder automates operational decisions, but its public proposition runs ahead of the demonstrated integration of those decisions. Replenishment remains organized around service targets and buffers; manufacturing and network optimization depend on configured objectives and constraints; newer agents investigate exceptions and compare scenarios. The company does not show how these mechanisms jointly value uncertain demand, supply, substitution, capacity, and cash across the portfolio. Shared data and faster orchestration leave that economic problem unresolved. (2, 3, 7, 16)
The strongest technical evidence belongs to identifiable components: published demand-density research, inspectable forecasting code, and an implementation of automated fulfillment using Enterprise Supply Planning. These establish capabilities within their scope. The weakest parts of the proposition concern the recurring cost of maintaining a heterogeneous suite, the productivity added by its latest agent layer, and the accountability behind broad autonomy and return-on-investment claims. Morrisons’ disclosures about the 2024 cyberattack also make recovery an observed supply chain issue, with consequences for availability and waste. (4, 6, 7, 20, 23)
Blue Yonder versus Lokad
Blue Yonder connects a large collection of planning and execution applications. Lokad concentrates on the recurring calculations that determine purchases, production orders, allocations, and other supply chain decisions. Its inventory optimization documentation describes customer-specific economic logic, probabilistic demand and lead times, and Supply Chain Scientists responsible for maintaining that logic.
Consider a retailer deciding whether to replenish a slow seller, use its shelf space for a substitute, or discount the remaining stock. A cheaper shipment may arrive after the season; a higher service target may consume cash better spent elsewhere. These choices require a monetary comparison of uncertain outcomes. Blue Yonder offers replenishment, assortment, pricing, fulfillment, and returns capabilities, but their common ownership and shared data do not establish one calculation that retains all these interactions. Its replenishment FAQ moves between category service targets and profit claims without explaining how conflicts between those objectives are resolved. (2, 9, 25)
Lokad’s design puts the decision calculation and its improvement at the center. When repeated recommendations are economically wrong, the task is to revise the assumptions, data interpretation, or available actions and test the revised logic. This perspective is developed in Introduction to Supply Chain and Holimization. Its practical value depends on the resulting decisions and the full cost of the service and software that maintain them.
There is a useful Blue Yonder comparison at this same operational level. Spinnaker SCA describes replacing a food distributor’s daily fulfillment overrides with rules embedded in Blue Yonder’s LP engine. The customer then maintained prioritization weights rather than repairing each plan. That is a better use of people than repetitive order review. The remaining limitation is how far the underlying formulation can evolve when a missed economic interaction cannot be expressed by adjusting those weights. (7)
What the decision engines actually do
Replenishment mixes economic language with service policies
Blue Yonder’s replenishment description identifies actionable outputs: orders to vendors and shipments to stores. It includes minimum quantities, pallet and truckload rules, multi-echelon replenishment, demand volatility, and lead-time variability. Its dynamic safety-stock description likewise adjusts buffers by SKU and node as demand and supply conditions change. These are material operational inputs. (2, 8)
The governing economic rule is less clear. The replenishment FAQ gives an example of 98% availability for high-velocity items and 90% for low-velocity items, then claims profit-centric decisions that can accept a stockout to avoid an expensive delivery. Those approaches can produce different orders. A velocity class cannot determine whether a shortage is costly: a slow-moving spare can stop a production line, while a fast seller may have a close substitute. The public description leaves the reconciliation between service policy and monetary consequences unspecified. (2)
SiteOne’s deployment illustrates the conventional workflow. Statistical demand forecasts feed distribution requirements planning; case quantities, vendor minimums, and lead times constrain replenishment; Order Optimization balances safety stock and purchase efficiency. SiteOne reports 97% in-stock availability and says it is still improving its measurement framework. That percentage describes stock availability, not forecast accuracy or net economic return. The article does not quantify the capital, markdowns, or total labor needed to sustain it. (19)
Probabilistic demand has an inspectable mechanism
Blue Yonder researchers published a method that predicts a negative-binomial demand distribution for each item, store, and day, estimating its mean and dispersion separately with Cyclic Boosting. The paper examines calibration and temporal confounding, and connects demand distributions to the costs of under- and over-ordering. Its example uses 100 products across ten Walmart stores, with pre-2016 training data and 2016 observations reserved for testing. It is an experiment on demand distributions, not a deployed evaluation of the complete Blue Yonder suite. (4)
Cyclic Boosting is a generalized additive approach that iteratively updates contributions from binned features. Its source code exposes the negative-binomial dispersion estimator and likelihood calculation. This makes a specific part of Blue Yonder’s forecasting work reproducible and technically discussable. The method’s interpretability concerns the contribution of forecasting features; it does not explain every downstream purchasing or production decision. (5, 6)
A distribution for each item’s demand is still short of the uncertainty needed for coupled decisions. Suppose two products depend on the same late shipment and can substitute for one another. Separate demand distributions and a lead-time-variability input do not specify their joint outcomes or which product should receive scarce stock. The reviewed technical material does not trace such dependencies into a common economic decision policy. Blue Yonder’s 2026 language about a probability corridor culminates in scenario comparisons and buffer adjustments, without exposing the probability model that selects actions. (4, 8, 17)
LP and MILP constrain what can improve
Blue Yonder’s network-optimization FAQ explicitly identifies mixed-integer linear programming. It describes minimizing total cost or maximizing service, then presents the resulting global optimum as the best network decision. Spinnaker’s food-distribution implementation identifies an LP engine in Enterprise Supply Planning, with service commitments, margins, lead times, customer classes, and perishability translated into fulfillment priorities. These are two documented applications, with different scopes. (3, 7)
An optimum establishes that the configured model has been solved. It does not establish that its formulation preserves the economics of the operation. Full-truck purchases, setup costs, substitution, uncertain receipts, and future corrective actions create discrete and nonlinear interactions. Linearization can discard them; retaining them with integer variables and scenarios adds computational burden. Repeatedly solving an updated forecast does not price the uncertainty that existed when the order became irreversible.
The more consequential test is whether a discovered failure can lead to a lasting improvement. If a replenishment recommendation repeatedly neglects shelf-space competition, adding another exception threshold leaves the defect in the decision model. A useful architecture must retain the competing assortment and replenishment choices, their uncertain consequences, and the relevant costs while continuing to run at operational scale. The public LP implementation documents adjustable priorities; it does not document this broader reformulation capability. (7)
Order promising has a narrower, observable role: choose fulfillment locations, allocate stock and resources, and estimate delivery dates using costs, margins, speed, and availability. That is a decision mechanism close to execution. Its public description does not extend the same monetary objective backward through assortment, purchasing, production, and returns. (9)
Technology history: autonomy promised repeatedly
Blue Yonder’s history contains both deployed automation and repeated expansion of the promise. Keeping those threads separate makes the current claims easier to assess.
| Period | Public commitment or delivery | What it establishes |
|---|---|---|
| 2017–2018 | Morrisons reported automated ordering across its stores; Blue Yonder’s case study attributes lower stock and improved availability to forecasting and replenishment. | Store ordering automation predates generative AI. The customer’s annual report corroborates deployment, but does not isolate a software-only financial return. (10, 11) |
| August 2018 | JDA’s Microsoft announcement promoted cognitive SaaS and an autonomous supply chain, with Luminate ControlTower as its first Azure application. | Cloud infrastructure and a control tower were offered as the route to much broader autonomy. (12) |
| March 2019 | JDA said Luminate could predict events, prescribe action, and act autonomously. The described Control Tower workflow used what-if analysis and guided choices; Demand Edge produced short-term probabilistic forecasts. | The specific workflow remained narrower than the headline autonomy claim. (13) |
| January 2024 | Blue Yonder announced its first interoperable solutions spanning planning, warehouse, transportation, and commerce, alongside Snowflake-backed data services. | Suite-wide interoperability was still being released years after the earlier end-to-end vision. (14) |
| 2026 | Blue Yonder promoted Agentic Ops, Orchestrator, MCP integration, third-party ML models, scenario generation, and guided response workflows. | The new interface and extension layer is identifiable; its incremental labor and economic results remain unquantified in this account. (16) |
The 2019 materials are particularly revealing. They describe task-specific prediction and decision support while announcing autonomous behavior across the supply chain. The autonomy claim overextended the demonstrated mechanism: the announcement supplied no operational account of unattended cross-functional decisions or residual human work. The relevant criticism is the unresolved gap between the advertised operating model and the concrete workflow. Morrisons’ earlier order automation does not validate that wider promise. (12, 13)
The 2024 release also included a Unified Commerce Simulator for comparing sourcing strategies against production data. Replay is useful for investigating a changed policy, but a digital twin that predicts what would happen under different decisions needs validation of those counterfactual outcomes. The release supplies no error analysis for those forecasts. Faster simulation can make a mistaken economic frame easier to explore without making it more accurate. (14)
Acquisitions continue to enlarge the engineering task. The Doddle announcement added returns capabilities; flexis added production and transportation planning; the August 2024 One Network acquisition added multi-enterprise collaboration for approximately $839 million. The latter announcement explicitly described integration into the Blue Yonder Platform as forthcoming. Current shared-platform and network descriptions show progress in connectivity, but do not document retirement of all older data models and application runtimes. (15, 25, 27, 28)
Software materials and the cost of integration
The product portfolio has several identifiable technical environments. The table keeps employer-described application stacks separate from skills requested for implementation work.
| Area | Publicly identified materials | Operational implication |
|---|---|---|
| Platform Data Cloud | Snowflake, common workflows, an in-memory cube, a shared database, and a Data Share App distributed through a private Snowflake Marketplace listing. | Data integration, analytical compute, and application execution have distinct costs and failure boundaries. (26) |
| SCPO implementation services | Job 261200, dated September 11, 2026, requires PL/SQL, ETL, Snowflake, Python scripting, integration design, and performance tuning. | Installed planning environments still require specialist implementation and maintenance. The role spans customer implementation, integration, and tuning. (20) |
| Warehouse returns processing | Job 262966 identifies TypeScript/JavaScript, Python and Java; React, Node.js, FastAPI and Spring Boot; MongoDB, PostgreSQL and Snowflake; Docker, Kubernetes, Azure/AWS and GitHub Actions. | Even this bounded product area has several persistence systems, runtimes, and deployment layers to maintain. (21) |
| Published forecasting library | Cyclic Boosting’s negative-binomial implementation uses NumPy, Numba and scikit-learn interfaces. | An inspectable forecasting component is available, independently of the proprietary application stack. (6) |
The warehouse-returns inventory is substantial. Multiple databases can serve different workloads, but someone must maintain consistent meanings for an item, its condition, location, and recoverable value across the relevant stores and services. A return routed to resale rather than liquidation has to propagate into availability, replenishment, and financial valuation. Shared access to data does not settle which application owns that choice or how conflicting updates are reconciled.
Blue Yonder advertises faster processing and a twentyfold cost reduction for services running on Snowflake, without identifying the comparative workload and cost boundary on the reviewed page. Such a figure cannot establish lower total ownership cost. The hosting bill also includes services outside that calculation; the customer’s bill includes migrations, integration changes, regression tests, and staff able to diagnose failures across products. (26)
The SCPO vacancy makes this labor explicit. It assigns consultants responsibility for custom integrations, data mappings, workflow and interface changes, performance tuning, and estimating work outside the standard solution. These are recurring software responsibilities. Customers need to know which party owns their custom logic after a release changes a schema, a solver, or an upstream ERP interpretation. The reviewed material does not show a common, portable representation of that logic across the suite. (20)
Productivity: count the work that remains
Morrisons’ automated store ordering is a more meaningful productivity reference than today’s agent branding. Its annual report described deployment across stores and lower costs; the vendor’s case study documents the previous dependence on manual ordering. The mechanism removed a repeated operational task. (10, 11)
Spinnaker’s ESP case describes another specific improvement: five daily optimization runs, substantially fewer overrides, and a final result sent to the warehouse with little intervention. That account identifies how work changed. It supplies no comparable headcount, maintenance hours, or net financial calculation, so it cannot establish the complete productivity gain. (7)
Core-Mark’s space-planning case exposes the work behind another improvement. Following an 18-month transformation, analysts maintained 525 reference planograms, refreshed quarterly, and adapted them to store fixtures and regional product libraries. A 70-store chain’s turnaround fell from two weeks to three days. Generative agents were still being evaluated for the next stage. This is evidence of faster planogram production and continued specialist maintenance, not a measured benefit of the prospective agents. (18)
The current Inventory Ops Agent is presented as a way to find data and plan issues and focus planners on consequential SKUs and orders. The August 2026 agent account adds root-cause analysis, replenishment settings, and scenario comparisons. These are recognizable planner tasks. Reproducing them through conversational agents can leave the work growing with the number of exceptions instead of improving the reusable calculation that produces decisions. (1, 16)
MCP integration therefore earns no independent productivity credit. Connecting a language model to an existing application is a small part of the engineering required to eliminate operational work. Blue Yonder needs to establish what its agents remove beyond existing replenishment and fulfillment automation, how often they require correction, and how much effort remains across customers, consultants, and vendor support.
Security, recovery, and economic accountability
Morrisons’ March 2025 trading update attributed disruption of stock accuracy, availability, waste, and forecasting to the Blue Yonder cyberattack. Its June update reported recovery following the November 2024 incident. These customer disclosures establish operational consequences; they do not identify the intrusion mechanism or implicate every Blue Yonder hosting environment. (23, 24)
Blue Yonder’s current security page describes air-gapped backups in separate Azure regions, with immutable and indelible backup data, network segmentation, and recovery testing. Those are relevant design measures because they address damage after compromise. The missing evidence concerns enforcement, tested recovery times, and the application coverage of the measures. Certification frameworks and a statement of security culture cannot answer those questions. (22)
The platform also advertises one authentication and authorization model. Consolidation can reduce inconsistent access controls, but it makes the limits of a compromised identity more important. The public description does not expose the permissions separating planning, execution, administration, and agent actions. The announcement of customer-controlled ML also leaves the containment of imported models unexplained. (16, 20, 25)
Economic accountability remains weaker than operational deployment evidence. The Morrisons case study attributes 2.6% same-store sales growth to its planning solution, without separating the contribution of the software from the retailer’s other commercial changes. SiteOne’s 97% in-stock figure measures a different outcome. Neither defines a baseline and return net of subscriptions, implementation, maintenance, and remaining labor. (11, 19)
The appropriate responsibility follows control: Blue Yonder owns its engines and hosted services; integrators often own custom mappings and formulations; customers own some policies and execution. A useful economic commitment would bound those obligations and identify who diagnoses and repairs an unprofitable recommendation. Public success stories do not provide that commitment. This gap matters more as a suite accumulates modules and partners, because a failed decision can cross several organizational boundaries before anyone accepts responsibility for changing its calculation.
Supply chain score
The scores are editorial judgments on the documented software and operating model. Each of the 25 criteria has equal weight.
Supply chain depth
| Criterion | Score |
|---|---|
| Economic framing | 6/10 |
| Decision end-state | 8/10 |
| Conceptual sharpness on supply chain | 6/10 |
| Freedom from obsolete doctrinal centerpieces | 3/10 |
| Robustness against KPI theater | 4/10 |
Average: 5.4/10.
Economic framing — 6/10. Fulfillment sourcing and the documented ESP implementation incorporate margins and costs. Replenishment also claims profit-based choices. The missing common objective becomes consequential when replenishment, assortment, service policies, and fulfillment disagree about where the next dollar should go. (2, 7, 9)
Decision end-state — 8/10. The suite produces executable replenishment and fulfillment decisions, with customer and implementation accounts showing repeated operational use. The principal limitation is the boundary between automated application outputs and the broader planning process, where scenario selection and exception investigation remain human work. (10, 16, 19)
Conceptual sharpness on supply chain — 6/10. The density-forecasting research explicitly connects uncertainty to ordering economics. Public product explanations are less consistent: the same replenishment narrative alternates between availability targets and profitability without exposing their arbitration. That ambiguity affects what a customer is actually configuring. (2, 4)
Freedom from obsolete doctrinal centerpieces — 3/10. Safety-stock policies, class-based service targets, consensus planning, and scenario review remain central in the documented offering. Automating their calculation preserves the limitations of those proxies when costs, substitution, and shortages vary more finely than the policy classes. (2, 8, 29)
Robustness against KPI theater — 4/10. Public cases identify operational metrics, but generally leave capital, total labor, and attribution incomplete. SiteOne’s availability result cannot settle profitability, while Morrisons’ same-store sales growth cannot be assigned wholly to a planning engine without a counterfactual. (11, 19)
Decision and optimization substance
| Criterion | Score |
|---|---|
| Probabilistic modeling depth | 6/10 |
| Distinctive optimization or ML substance | 7/10 |
| Real-world constraint handling | 7/10 |
| Decision automation and labor productivity | 6/10 |
| Resilience under real operational complexity | 4/10 |
Average: 6.0/10.
Probabilistic modeling depth — 6/10. Published distribution modeling and calibration go materially beyond a point forecast with a confidence band. The supply side has a public variability claim, but the decision chain lacks an inspectable joint model that connects uncertain receipts and dependent demand to coupled actions. (4, 8)
Distinctive optimization or ML substance — 7/10. Cyclic Boosting provides a named method, equations, and executable implementation. This is stronger evidence than an AI label. Its contribution concerns forecasting; the proprietary optimization applications remain much less inspectable, and generic MILP capability does not establish a distinctive economic method. (3, 5, 6)
Real-world constraint handling — 7/10. Replenishment and fulfillment address physical ordering and distribution restrictions, with implementation evidence for a constrained food network. The limitation is the treatment of interactions that cannot be reduced economically to a fixed rule or linear priority. (2, 7)
Decision automation and labor productivity — 6/10. Operational ordering automation is established, and Core-Mark supplies a concrete reduction in turnaround time. The current agent material provides no comparable accounting of work eliminated, residual reviews, or maintenance across the complete process. (10, 16, 18)
Resilience under real operational complexity — 4/10. Adjustable priorities and replay facilities support changes within an existing formulation. Public evidence is thinner on retaining newly discovered economic interactions through reformulation and controlled deployment. A faster solve cannot repair a decision space that excludes the necessary action. (7, 14)
Product and architecture integrity
| Criterion | Score |
|---|---|
| Architectural coherence | 4/10 |
| System-boundary clarity | 6/10 |
| Security seriousness | 3/10 |
| Software parsimony and ownership cost | 3/10 |
| Compatibility with programmatic and agent-assisted operations | 5/10 |
Average: 4.2/10.
Architectural coherence — 4/10. The common data layer and newer interoperability provide an integration direction. Employer descriptions still expose materially different application and implementation environments. A consistent data view is useful, but the public architecture does not settle ownership of overlapping calculations or retirement of legacy dependencies. (14, 20, 21)
System-boundary clarity — 6/10. Planning, order execution, warehouse operations, and the trading-partner network have recognizable roles. Boundaries become less clear under the claim of one intelligent platform: the public materials do not trace an order across authoritative state, decision ownership, and reconciliation between all participating applications. (15, 25)
Security seriousness — 3/10. The documented recovery design addresses a relevant failure mode. Morrisons’ experience shows the economic damage that actually reached a customer; public validation of recovery limits and present application coverage remains insufficient. Compliance claims add no evidence about the blast radius of a compromised account or service. (22, 23, 24)
Software parsimony and ownership cost — 3/10. The returns stack and SCPO services responsibilities expose substantial infrastructure and specialist maintenance. The advertised cloud cost reduction lacks a workload and full-cost boundary. Customers cannot infer a low lifetime bill from a narrower compute comparison. (20, 21, 26)
Compatibility with programmatic and agent-assisted operations — 5/10. APIs, data egress, and automated engines provide programmatic access. The latest agent descriptions emphasize investigation and guided workflows more than authoring, testing, and safely improving reusable decision logic. MCP connectivity alone does not close that gap. (16, 25)
Technical transparency
| Criterion | Score |
|---|---|
| Public technical documentation | 6/10 |
| Inspectability without vendor mediation | 5/10 |
| Portability and lock-in visibility | 4/10 |
| Implementation-method transparency | 5/10 |
| Evidence density behind technical claims | 5/10 |
Average: 5.0/10.
Public technical documentation — 6/10. The research and forecasting source code support detailed scrutiny of a numerical component. Comparable public specifications for the suite’s stochastic decision policy, solver formulations, runtime permissions, and cross-product semantics are missing from the reviewed material. (4, 5, 6)
Inspectability without vendor mediation — 5/10. An engineer can inspect and run the published forecasting implementation. Assessing the proprietary application’s end-to-end calculation still depends on Blue Yonder or an implementation partner; feature pages do not make that calculation reproducible. (6, 7, 25)
Portability and lock-in visibility — 4/10. The platform advertises data egress and a common API set. Exporting data is easier to establish than transferring custom planning logic, configured priorities, and integrations to another runtime. The implementation responsibilities show why those artifacts can become an expensive dependency. (20, 25)
Implementation-method transparency — 5/10. The ESP and Core-Mark accounts explain specific changes to business work, while the SCPO role reveals specialist tasks. They leave testing effort, upgrade responsibilities, and long-term maintenance costs largely unquantified. (7, 18, 20)
Evidence density behind technical claims — 5/10. Published methods and named customer workflows support selected capabilities. Broad claims of autonomous, probabilistic, end-to-end decision-making still lack the integrated model and outcome evidence needed to evaluate them as a single proposition. (4, 16, 17)
Vendor seriousness
| Criterion | Score |
|---|---|
| Technical seriousness of public communication | 4/10 |
| Technology roadmap integrity | 4/10 |
| Conceptual sharpness | 5/10 |
| Economic accountability | 3/10 |
| Defensibility in an agentic-software world | 6/10 |
Average: 4.4/10.
Technical seriousness of public communication — 4/10. Technical publications coexist with claims that a mathematical optimum identifies the best business decision and that a data-cloud implementation dramatically reduces costs without a defined comparison. The latter claims hide the assumptions on which their economic relevance depends. (3, 4, 26)
Technology roadmap integrity — 4/10. Deployed ordering automation provides historical delivery evidence. The broader autonomy claim persists from 2018–2019 through the agent era, while 2024 introduced the first suite-wide interoperable solutions. The current communication does not reconcile those successive boundaries or quantify the promised progression. (10, 12, 13, 14, 16)
Conceptual sharpness — 5/10. The forecasting research distinguishes uncertainty, costs, and prediction quality. Product communication blurs scenario planning, probabilistic decision-making, and autonomy. That imprecision makes it difficult to identify the actual engine being purchased behind the latest vocabulary. (4, 17)
Economic accountability — 3/10. Cases report favorable outcomes without a complete net-return calculation, bounded customer obligations, or published remediation responsibility for poor recommendations. The documented customer disruption reinforces the need to include operational losses and recovery in that economic account. (11, 19, 23)
Defensibility in an agentic-software world — 6/10. Operational applications, established integrations, and inspectable forecasting methods are substantial assets. Their value must outweigh the cost of a large maintained portfolio as agents make application access and routine analysis cheaper. The newer connector and assistant layer supplies little evidence of an additional advantage in decision quality or improvement productivity. (6, 15, 16, 20)
Sources
All sources were accessed on November 4, 2026. Publication dates are listed where available.
[1] Demand and Supply Planning
Demand and Supply Planning. undated. Current planning functions and Inventory Ops Agent workflow.
[2] What is Blue Yonder Replenishment for Retail?
What is Blue Yonder Replenishment for Retail?. undated. Ordering outputs, logistics constraints, service targets, and profit claims.
[3] What is Network Optimization?
What is Network Optimization?. undated. MILP formulation, objectives, and global-optimum claims.
[4] Demand Forecasting of Individual Probability Density Functions with Machine Learning
Demand Forecasting of Individual Probability Density Functions with Machine Learning. 2021-07-22 (v3). Negative-binomial demand modeling, calibration, and the Walmart experiment; version 3.
[5] Cyclic Boosting: an explainable supervised machine learning algorithm
Cyclic Boosting: an explainable supervised machine learning algorithm. 2020-02-09 (arXiv submission). Cyclic Boosting formulation and interpretability.
[6] Cyclic Boosting negative-binomial regressor, commit 7e7cee2
Cyclic Boosting negative-binomial regressor, commit 7e7cee2. accessed 2026-11-04. Versioned source code for the negative-binomial dispersion estimator.
[7] Optimized Order Promising for a Leading Food Distributor
Optimized Order Promising for a Leading Food Distributor. undated. Implementer account of ESP-based fulfillment, LP priorities, and recurring operation.
[8] What is Dynamic Safety Stock?
What is Dynamic Safety Stock?. undated. Demand and lead-time variability in SKU/node buffer sizing.
[9] Order Promising and Optimization
Order Promising and Optimization. undated. Fulfillment sourcing inputs and order-promising functions.
[10] Morrisons Annual Report and Financial Statements 2017/18
Morrisons Annual Report and Financial Statements 2017/18. 2018. Customer disclosure of automated ordering and operating benefits.
[11] Morrisons Puts the Customer at the Heart of Every Decision
Morrisons Puts the Customer at the Heart of Every Decision. undated. Vendor account of Morrisons replenishment and attributed business results.
[12] JDA Partners with Microsoft to Power Data-Driven Digital Transformations in the Cloud
JDA Partners with Microsoft to Power Data-Driven Digital Transformations in the Cloud. 2018-08-01. JDA-issued announcement of the Azure partnership and autonomy vision, distributed by WebWire.
[13] JDA Showcases Powerful Luminate Portfolio Built on Microsoft Azure at Hannover Messe
JDA Showcases Powerful Luminate Portfolio Built on Microsoft Azure at Hannover Messe. 2019-03-15. JDA-issued account of Control Tower and Demand Edge, published by IT Supply Chain.
[14] Blue Yonder Launches Interoperable Solutions To Unlock Performance and Build Supply Chain Resilience
Blue Yonder Launches Interoperable Solutions To Unlock Performance and Build Supply Chain Resilience. 2024-01-11. Interoperability release, Snowflake data services, and commerce simulation.
[15] Blue Yonder Acquires One Network Enterprises
Blue Yonder Acquires One Network Enterprises. 2024-08-01. Acquisition close, network scope, and planned platform integration.
[16] Blue Yonder recognized again as a Leader in the Nucleus Research 2026 Enterprise Supply Chain Planning Technology Value Matrix
Blue Yonder recognized again as a Leader in the Nucleus Research 2026 Enterprise Supply Chain Planning Technology Value Matrix. 2026-08-19. Agentic Ops, Orchestrator, MCP, customer models, and guided workflows.
[17] Why your supply chain plan is already obsolete
Why your supply chain plan is already obsolete. 2026-02-11. Probability-corridor positioning, scenario reruns, and buffer adjustment.
[18] Core-Mark boosts planogram coverage by 40% with Blue Yonder
Core-Mark boosts planogram coverage by 40% with Blue Yonder. undated. Planogram maintenance, rollout duration, turnaround improvement, and prospective agents.
[19] SiteOne Landscape Supply transforms forecasting and order management
SiteOne Landscape Supply transforms forecasting and order management. undated. Deployed forecast/replenishment workflow, in-stock outcome, and measurement limitations.
[20] Enterprise Technical Architect: SCPO, Planning (job 261200)
Enterprise Technical Architect: SCPO, Planning (job 261200). 2026-09-11. Employer-authored SCPO implementation responsibilities and technical requirements.
[21] Sr Software Engineer: ReactJS, TypeScript and SQL/NoSQL (job 262966)
Sr Software Engineer: ReactJS, TypeScript and SQL/NoSQL (job 262966). Undated; accessed November 4, 2026. Employer-authored warehouse-returns development stack.
Security at Blue Yonder. undated. Current backup, recovery, access, and compliance statements.
[23] Morrisons trading update for Q1 2024/25
Morrisons trading update for Q1 2024/25. 2025-03-26. Customer disclosure of operational damage following the Blue Yonder cyberattack.
[24] Morrisons trading update for Q2 2024/25
Morrisons trading update for Q2 2024/25. 2025-06-19. Customer follow-up describing recovery after the November 2024 attack.
Blue Yonder Platform. undated. Platform scope, extension framework, access model, APIs, and egress.
[26] Data cloud and supply chain platform
Data cloud and supply chain platform. undated. Data-cloud architecture, sharing, and cost/performance claims.
[27] Blue Yonder Acquires flexis
Blue Yonder Acquires flexis. 2024-02-08. Acquisition of manufacturing and transportation-planning software.
[28] Blue Yonder Announces Intent to Acquire Doddle
Blue Yonder Announces Intent to Acquire Doddle. 2023-10-12. Original returns-technology acquisition announcement.
[29] Supply Chain Planning FAQs
Supply Chain Planning FAQs. undated. Consensus planning and the relationship between planning functions.